Gene expression biomarker to predict estrogen receptor activity
High-throughput transcriptomics (HTTr) has the potential to support efforts to reduce or replace some animal tests. EPA and NIEHS scientists developed a computational approach utilizing MCF-7 breast cancer cells and a biomarker assessing expression of 46 genes to predict estrogen receptor activity after chemical exposure (Ryan et al. 2016). To further explore the utility of this model, they investigated whether it could identify estrogen receptor activities of chemicals examined by Endocrine Disruptor Screening Program (EDSP) Tier 1 screening assays (Corton et al. 2022). For the 62 chemicals examined, the estrogen receptor biomarker model accuracy was 97% for in vitro reference chemicals and at least 76% for in vivo assays. These accuracies were similar or slightly better for the same chemicals than those of a previously described ToxCast estrogen receptor model based on 18 in vitro assays. These results indicate that the HTTr biomarker model can correctly identify active and inactive estrogen receptor reference chemicals, and is potentially useful to rapidly identify chemicals with potential estrogen receptor bioactivities for additional screening and testing.